Justification Review

October 2003 Report No. 03-52 Economic Self-Sufficiency Performance Mixed; Food Stamp Improvements Could Yield Federal Bonuses at a glance This report reviews the Economic Self- The Economic Self-Sufficiency (ESS) Program helps low-income individuals and families become self-sufficient. While the program has improved its timeliness in processing applications for services, its continuing high food stamp error rates have resulted in Florida receiving federal penalties totaling over $2.2 million for federal Fiscal Years 2001 and 2002. Improved performance could qualify Florida for federal bonus funding beginning in federal Fiscal Year 2003. While the program s efforts to prevent and recover overpayments have yielded an overall positive return on investment, monies collected in overpayments have not covered costs for the past two years. Sufficiency (ESS) program administered by the Department of Children and Families. The report evaluates program performance based on timeliness, accuracy, and return on investment, assesses program efforts to avoid costs through fraud prevention and recovery of benefits improperly paid to clients, and makes recommendations for improving program operations and enabling the state to receive federal performance funding by reducing food stamp error rates. Background Scope Florida s Economic Self-Sufficiency (ESS) Section 11.513, Florida Statutes, directs the Office of Program Policy Analysis and Government Accountability to conduct justification reviews of programs operating under a performance-based program budget. Justification reviews assess agency performance measures and standards, evaluate program performance, and identify policy alternatives for improving services and 1 reducing costs. Program assists low-income individuals and their families to become and remain stable and self-sufficient. In Fiscal Year 2002-03, the program provided services to nearly 2.8 million Florida citizens. 2 Client Services. The ESS Program helps lowincome Florida citizens who meet eligibility requirements access a variety of services intended to assist them to meet their basic 1 Chapter 92-249, Laws of Florida (see Appendix A for statutory requirements). 2 Florida s TANF program ranks twelfth nationally and its Food Stamps and Medicaid programs rank fourth nationally in terms of the number of individuals served by these programs. Office of Program Policy Analysis and Government Accountability an office of the Florida Legislature

needs. These services can be categorized as described below. Food Stamps. The program determines eligibility for and issues food stamps to lowincome individuals who qualify. Individuals receiving cash benefits from Temporary Assistance for Needy Families (TANF), Supplemental Security Income (SSI), Social Security Disability Insurance (SSDI), and refugee assistance payments automatically qualify to receive food stamps. Other persons who meet established criteria can also qualify to receive food stamps. Food stamp benefits are 100% funded by the federal government. The state and federal governments equally share administrative costs. On average, in 2002-03, ESS provided food stamps to 1,024,692 individuals per month. The program authorized $964,194,474 in annual food stamp benefits, averaging $924.90 per person. Cash Assistance. ESS determines eligibility for and provides cash assistance to individuals who qualify for TANF and refugee assistance. TANF provides cash benefits to certain low-income individuals including adults with dependent children and children residing in foster care. ESS also provides cash assistance to refugee adults and families. Cash assistance to refugees is 100% federally funded while the state and federal governments share the cost of TANF cash assistance. On average, ESS provided cash assistance to 129,402 individuals during each month of 2002-03. For that same year, cash assistance benefits totaled $226,582,830 and averaged $686.67 per person. Medicaid. Although administered by the Agency for Health Care Administration, the ESS program determines Medicaid eligibility for Florida citizens. Like food stamps, individuals who receive cash benefits from TANF, SSI, SSDI, and refugee assistance automatically qualify for Medicaid. In addition, Medicaid provides health care services to other persons who meet established income and assets criteria. Medicaid funding comes from both the state 2 and federal governments. In 2002-03, ESS determined Medicaid eligibility for 2,304,769 individuals. State programs. ESS determines financial eligibility for two services that are primarily state funded. 3 Optional state supplementation (OSS) is a cash assistance program to help pay for community alternative living arrangements and prevent placement in an institution. 4 Home care for disabled adults provides a monthly payment to persons who provide home care for eligible disabled adults less than 65 years of age. On average, ESS provided 9,875 individuals with OSS assistance during each month of 2002-03. Total expenditures for OSS assistance were $23,113,061 and averaged $1,707 per individual. Other Services. Besides the major public assistance services, ESS provides emergency financial housing assistance and funds to local communities to assist them in providing services to the homeless. ESS also awards grants to individuals and communities affected by natural disasters. The total value of these awards for Fiscal Year 2002-03 was $1,459,072. Organization. The Department of Children and Families plans, administers, and delivers most ESS services through 13 district offices and one regional office. District and regional offices are responsible for ensuring that services are delivered in accordance with state and federal laws and for coordinating ESS services with other public or private agencies that offer services for clients. (See Exhibit 1.) 3 While historically 100% state funded, a 2000-01 amendment to the Medicaid state plan allows some federal funds to support assistive care costs in the OSS program. 4 These alternative arrangements are assisted living facilities or adult family care homes.

Exhibit 1 ESS Provides Services to Low-Income Persons Through 13 Districts and One Region 1 2 1 - Escambia, Okaloosa, Santa Rosa, Walton 2 - Bay, Calhoun, Franklin, Gadsden, Gulf, Holmes, Jackson, Jefferson, Leon, Liberty, Madison, Taylor, Wakulla, Washington 3 - Alachua, Bradford, Columbia, Dixie, Gilchrist, Hamilton, Lafayette, Levy, Putnam, Suwannee, Union 4 - Baker, Clay, Duval, Nassau, St. Johns Suncoast Region DeSoto, Hillsborough, Manatee, Pasco, Pinellas, and Sarasota 7 - Brevard, Orange, Osceola, Seminole 8 - Charlotte, Collier, Glades, Hendry, Lee 9 - Palm Beach 10 - Broward 11 - Dade, Monroe 12 - Flagler, Volusia 13 - Citrus, Hernando, Lake, Marion, Sumter 14 - Hardee, Highlands, Polk 15 - Indian River, Martin, Okeechobee, St. Lucie 3 Source: Department of Children and Families. 4 13 12 14 7 15 8 9 10 11 The ESS central office located in Tallahassee is responsible for statewide program oversight and administration. Central office personnel develop policy, provide technical assistance and training, develop and manage the annual budget, and identify and recover benefits that have been overpaid. Central office staff also monitors the effectiveness and efficiency of program operations and develops and maintains ESS information systems. Resources. For Fiscal Year 2003-04, the Legislature appropriated over $661 million to operate the ESS Program. The program uses 93.9% of its budget to directly assist low-income individuals. As shown in Exhibit 2, the program uses these funds to determine whether clients are eligible for services and to pay eligible clients benefits in the form of cash assistance. The program uses the remaining funds ($40.4 million, or 6.1%) to pay for administrative and oversight functions such as ensuring compliance with federal and state laws and recovering monies spent in error. 3 Exhibit 2 Most of the Program s $661 Million Budget Pays for Cash Assistance Benefits and Eligibility Determination and Case Management Program Oversight and Administration 6% Eligibility and Case Management 43% Source: LAS/PBS system, Fiscal Year 2003-04. Cash Assistance 51% For Fiscal Year 2003-04, the ESS program has 7,672 authorized full-time staff positions. The majority of these positions are district personnel responsible for determining and monitoring eligibility for economic self-sufficiency services. Findings The ESS Program has improved timeliness of processing applications A major goal of the Economic Self-Sufficiency (ESS) Program is to ensure that clients obtain assistance in a timely manner. The federal government requires states to process food stamp applications within 30 days, with deadlines of 45 and 90 days for processing applications for the other major assistance programs. 5 The Legislature established two timeliness standards as part of the program s performancebased budgeting (PB²) process. The first measure required the program to process 98% of all applications for the major assistance groups (food stamps, TANF, and Medicaid) within mandated timeframes in 2000-01, increasing the performance expectation to 99% 5 Processing applications includes interviewing applicants, collecting and verifying applicant provided information, and using verified information to determine the type and amount of benefits for which applicants qualify.

for 2001-02 and 2002-03. 6 The second standard requires the program to process 98% of Optional State Supplementation (OSS) applications within 90 days. The program improved timeliness for processing applications for the major assistance groups in 2001-02 and 2002-03. As shown in Exhibit 3, ESS has continued to improve overall timeliness. In Fiscal Year 2002-03, ESS essentially met the PB² standard, processing 98.97% of its major assistance group s applications within guidelines. This represents an improvement over its 2000-01 performance, when it processed 95.8% of applications within the deadline. During this two-year period, timeliness improved for each major assistance group, including an improvement in the timely processing of food stamp applications from The department attributes improved timeliness to enhancing the FLORIDA system and providing increased technical assistance to its field staff. 7 Over the last two years, the department enhanced the tracking capabilities of the FLORIDA system enabling Web technology to provide daily updates on whether timeliness problems are associated with scheduling appointments, determining eligibility, or processing pending claims. ESS public assistance workers and supervisors use this information to target actions for correcting local service center problems. To assist in this effort, central office employees have provided districts technical assistance on how to use these data to diagnose deficiencies and develop actions to improve performance. 93.5% to 98.3%. 7 The ESS Program uses the Florida On-line Recipient Integrated Data Access (FLORIDA) information system to track client eligibility and payments. 6 The Legislature has increased the performance standard for overall timeliness to 100% for 2003-04. Exhibit 3 ESS Timeliness for Processing Applications for Major Assistance Groups Continued to Improve in Fiscal Year 2002-03 Percentage Meeting Timeframe Mandated Timeframe 2000-01 2001-02 2002-03 Standard (2002-03) Major Assistance Groups Varies 95.8% 98.7% 99.0% 1 99% Food Stamps 30 days 2 93.5% 98.0% 98.3% 99% TANF Cash Assistance 45 days 98.5% 99.6% 99.6% 99% Medicaid, Children, Non-Disabled Adults 45 days 97.1% 99.1% 99.4% 99% Medicaid, Disabled Adults 90 days 98.3% 99.5% 99.6% 99% 1 The actual calculation was 98.97%of applications processed timely. 2 Most Food Stamp applications must be processed within 30 days, but emergency Food Stamp applications must be processed within 7 days. Source: Department of Children and Families. 4

The program improved timeliness for processing Optional State Supplementation (OSS) applications in 2002-03. The Legislature requires that OSS applications be processed within 90 days. As shown in Exhibit 4, the program processed 95% of its OSS applications within this time frame in 2002-03. While falling short of meeting the 98% standard, the program has improved its performance since 2000-01, when it processed only 75.6% of applications within the 90-day timeframe. Exhibit 4 The ESS Program Processed 95% of OSS Applications Within 90 Days During 2002-03, Improving From 75.6% in 2000-01 75.6% OSS Actual Performance Fiscal Year 2000-01 90.8% OSS Actual Performance Fiscal Year 2001-02 Source: Department of Children and Families. 95.0% OSS Actual Performance Fiscal Year 2002-03 The program s low performance in Fiscal Year 2000-01 was tied to a rapid increase in workload. Applications increased in anticipation of a Medicaid waiver that raised awareness of available benefits and lowered income eligibility criteria. As a result, between Fiscal Years 1999-2000 and 2000-01, the average number of applications processed increased 2.5% each month and the average caseload increased 11.9% each month. According to department managers, this increases workload resulted in OSS case workers processing only 75.6% of applications within 90 days. After the Medicaid waiver was implemented in September 2001, a large number of OSS cases were moved under Medicaid, decreasing OSS s 5 existing caseload by 8%. Over the next two Fiscal Years (2001-02 and 2002-03), the average number of applications processed each month decreased 10.6% and 3.5%, respectively while the average monthly caseload decreased by 3.7% each year. As workload decreased, program performance continued to improve. Accurately determining benefit eligibility and payments continues to be a challenge The ESS Program has not met state and federal standards for accuracy in determining benefit levels, and its continuing high food stamp error rates have resulted in Florida receiving federal penalties totaling over $2.2 million for federal Fiscal Years 2001 and 2002. 8 Program administrators attribute this problem to increased applications, unstable caseloads, high turnover rates for public assistance workers, and frequent changes in policy. The department no longer reports TANF accuracy; poor food stamp accuracy resulted in federal sanctions and missed funding opportunities Accurately determining how much TANF and food stamp benefits clients are eligible to receive is a critical program responsibility. Achieving a high level of accuracy is important to ensure that public funds are appropriately used and clients receive the proper benefits. The federal government may impose sanctions for high food stamp error rates and offers financial incentives for high or improved performance. The Legislature established accuracy performance standards for both TANF and food stamps as part of the PB 2 process. In addition, the USDA requires states to report food stamp error rates, the inverse of accuracy rates, and to calculate rates using federally approved sampling methods. 9 The USDA validates and 8 A federal fiscal year runs from October 1 through September 30. For example Fiscal Year 2003 runs from October 1, 2002, through September 30, 2003. 9 The US Department of Health and Human Services, federal

makes adjustments to error rates at the end of each federal fiscal year. 10 ESS stopped tracking and reporting TANF accuracy in 2001, contrary to state law. Although the Legislature requires the department to report TANF accuracy as a PB 2 measure, the department ceased monitoring these rates in October 2001. The department reported a 6.23% error rate for federal Fiscal Year 2000-01, the most recent year for which this data is available. While a lower error rate than in prior years, this rate indicated that the department overpaid clients by $13.6 million in that year. 11 As there is no ongoing monitoring, it is unknown if this error rate has improved or worsened since that date. While aware of the state requirement, program administrators stated that the department stopped tracking and reporting TANF payment accuracy because the federal government no longer requires the measure and the program used the staff formerly dedicated to this function for other tasks. However, the department s failure to report TANF accuracy hinders the Legislature s ability to ensure that TANF block grant dollars are spent wisely and that monies will be available to meet the needs of Florida s citizens. To comply with state law as well as to ensure fiscal responsibility, the department should track TANF payment accuracy rates, set annual improvement targets, and report results to the Legislature. Food stamp error rates exceeding the national average resulted in federal sanctions in federal Fiscal Years 2001 and 2002. As shown in Exhibit 5, Florida s food stamp error rate has exceeded the national average in four of the last five federal fiscal years. During this same time period, Florida s error rates were higher than two-thirds of the other states and U.S. territories administrator of the TANF program, has not required states to monitor and report error rates since welfare reform in 1996. However, until October 2001, the department continued to track TANF error rates using the federally approved methodology employed prior to 1996. 10 The USDA makes statistical adjustments to error rates for states with unusually high refugee and working poor caseloads. 11 TANF error rates include only errors that resulted in overpayments to clients, while food stamp error rates include both over and underpayments to clients. 6 for four of the five years. In general, Florida has performed poorly when compared to other large states. Exhibit 5 Florida s Food Stamp Error Rate Exceeded the National Average in Four of the Last Five Federal Fiscal Years 12.94% 10.70% 9.43% 9.40% 9.86% U.S. Average Florida 9.80% 9.61% 8.91% 8.66% 8.26% 1998 1999 2000 2001 2002 Source: USDA Food Stamp Program Accountability Division, Quality Control Branch. As the administrator of the federal food stamp program, ESS is expected to show good stewardship by minimizing error associated with paying benefits. However, in federal Fiscal Year 2002, Florida s error rate of 9.61% resulted in ESS overpaying clients $65.2 million in excess benefits and underpaying $19.2 million in benefits to eligible clients. 12 Because Florida s error rates exceeded the national average, the USDA has imposed financial penalties totaling more than $2.2 million against the state over the past two federal fiscal years. 13 The department is investing the penalty amounts to support activities designed to reduce error rates. In June 2003, the USDA offered the state a settlement agreement that allows the department to invest the 2002 penalty ($1.7 million) over federal Fiscal Years 2004 and 2005. The USDA will waive half 12 The overpayment error rate was 7.42% and the underpayment error rate was 2.19%. 13 The penalty for federal Fiscal Year 2001 was $541,466; the penalty for federal Fiscal Year 2002 was $1,715,327.

of the penalty amount if Florida achieves error rates at or below the national average in those years. Accomplishing this would, in effect, return nearly $860,000 to the state. In addition to penalties, Florida s high error rates prevented the state from receiving enhanced federal funding. Until federal Fiscal Year 2003, states that achieved error rates below 6% could earn up to 10% of their food stamp administrative costs in enhanced funding. Specifically, the USDA gave states an additional 1% of their food stamp administrative costs for each 0.1% reduction in error rate below 6%. Florida could have earned between $1.6 million and $15.6 million in federal Fiscal Year 2002 if it had reduced its food stamp error rate below 6%. 14 As shown in Appendix B, 16 states as well as the Virgin Islands earned incentive dollars at least once during the past five federal fiscal years. Texas, South Dakota, and Wyoming earned incentive funding for each of these five years, with Texas earning a total of $135 million during this time. Under the 2002 federal Farm Bill, Florida has new opportunities to earn bonuses for high or improved performance. By improving its food stamp accuracy rate, Florida can compete for a share of $48 million that the USDA will annually award to states with high or improved performance beginning in federal Fiscal Year 2002-03. 15 The USDA will award performance bonuses for federal Fiscal Year 2003 in these four areas. Application processing timeliness. $6 million to be divided among the six states with the highest percentage of applications processed within 30 days. Payment accuracy rate. $24 million to be divided among the seven states with the lowest and the three states with the most 14 We based estimates of potential incentive dollars that Florida could have earned using food stamp expenditures for federal Fiscal Year 2002. 15 Performance bonus awards were approved in Title IV of the Farm Bill (HR 2646), also known as the Food Stamp Reauthorization Act of 2002. 7 improved percentage of payments with the incorrect benefit amount. Negative error rate. $6 million to be divided among the four states with the lowest and the two states with the most improved percentage of cases where individuals met eligibility requirements but were denied benefits. Participation rate. $12 million to be divided among the four states with the highest and the four states with the most improved percentage of people below the poverty level who receive food stamps. The USDA will determine awards based on high performance or absolute improvement and in proportion to states caseloads, creating a significant opportunity for a large state like Florida that historically has had high error rates, but is performing well on other measures. 16 Almost half of food stamp errors are due to public assistance workers applying policies incorrectly or failing to take appropriate actions. Program administrators cite increased applications, high turnover rates for public assistance workers, and frequent changes in policy as reasons for the program s high error rates. The department s inspector general s office annually samples food stamp claims to determine error rates and uses these samples to identify the source and nature of inaccurate payments. 17 The office reports both the dollar value of errors and the percentage of cases with errors. As illustrated in Exhibit 6, slightly under half (44%) of the state s food stamp payment errors (in terms of dollar value) were due to errors made by public assistance workers, while 56% of errors were attributable to clients in federal Fiscal Year 2002. 18 16 Under the absolute improvement provision, a 2.8% improvement in error rate (e.g., if Florida s error rate decreased from 9.8% to 7%) would be given more weight than a state with a 2% improvement, even if the overall error rate for that state was lower (e.g., a change from 7% to 5% error rate). 17 Until October 2001, the inspector general conducted similar reviews of TANF claims, but discontinued these reviews when the department decided to stop monitoring TANF accuracy rates. 18 Based on the last available analysis of TANF errors (federal Fiscal Year 2001), public assistance workers were responsible for a greater proportion of errors (53.4%) than clients (46.6%).

Exhibit 6 Program Error Accounted for 44% of the Food Stamp Payment Errors in Federal Fiscal Year 2002 Program Error 44% Client Error 56% Source: Department of Children and Families, Office of Inspector General, Quality Control Annual Report, Food Stamp, federal Fiscal Year 2001-02. Historically, the most common types of errors made by public assistance workers have been similar for both TANF and food stamps. The majority of these errors are typically due to pubic assistance workers failing to act on information provided by clients or applying policies correctly. For example, a public assistance worker may fail to update the FLORIDA system to reflect a client s increased income, so the client receives more food stamps than he or she is entitled. The department is taking several steps to reduce food stamp error rates. We believe these actions, which have been successfully used in other states, hold potential for improving Florida s error rate problem. However, we believe these steps should be supplemented with increased training for public assistance workers that targets common problems. Piloting change units. Florida is piloting change units in Miami-Dade, the Suncoast Region (which includes DeSoto, Hillsborough, Manatee, Pasco, Pinellas, and Sarasota counties), and Orange County. The Suncoast Region change unit opened in October 2001; the Miami-Dade and Orange County units opened in late 2002. 8 Change units are staffed with personnel dedicated to updating client eligibility information such as changes in client income. Because these units are not responsible for processing new applications, employees can focus on ensuring that client eligibility information is current and benefit levels are accurately determined. The change unit model has been successfully used in other states. For example, Texas, which has consistently achieved a low error rate, has operated change units since 1997. Using federal options and waivers. Florida is using several federal program options and waivers and is pursuing others to simplify program administration. These options and waivers can help improve service delivery and decrease errors by providing states flexibility through simplifying rules, regulations and practices associated with determining and recertifying food stamp eligibility. 19 Currently the department uses several USDA options and waivers and is pursuing others. For example, Florida is pursuing a federal waiver called the Combined Outreach and Application Project, which would allow the state to merge the food stamp and TANF application process with the federal SSI application process. If this waiver is approved, any applicants inappropriately approved for benefits are not considered an error attributable to the state since the SSI eligibility was federally approved. 20 Increasing local accountability. The department is working to increase district accountability for error rates. Texas administrators identified this practice as a key factor in their ability to decrease food stamp payment errors. Texas establishes local performance benchmarks and then links managers employment to continuously 19 Options only require states to submit a letter of intent to the USDA to implement the option. Waivers require formal applications that outline the state s need and expected benefits of the requested change and must be specifically approved by the USDA. 20 See Appendix C for additional information related to federal food stamp options and waivers.

improving or meeting these standards. Florida has developed a case sampling methodology which will be used to establish local benchmarks. This methodology is currently being reviewed by the USDA. To fully implement this initiative, the department will need to specify the personnel at the district or regional level who will be held responsible for local performance. Updating the FLORIDA System. Finally, the department plans to streamline its eligibility determination process by using web-enabled data entry in its FLORIDA system which will include pull-down menus. These enhancements are intended to reduce errors by accommodating new policies and reducing training needs. 21 While the department originally planned to implement these changes in Fiscal Year 2001-02, it has delayed implementation indefinitely until funding becomes available. While the department s plans to reduce food stamp error rates show promise, they should be supplemented by targeted caseworker training. The department should use the results of its annual quality control reviews to identify common errors made by staff and target training for public assistance workers that highlights the policies and procedures that quality control reviews identify as commonly misapplied. In addition, the department should evaluate error reduction activities to determine whether to adjust or expand efforts. The department should annually report to the Legislature, with its first report due by December 31, 2003, on department efforts to evaluate these activities and their effect on reducing error rates. Fraud prevention and benefit recovery yield positive returns; recovery backlog cases have been reduced To ensure that public assistance dollars are spent appropriately, the ESS Program seeks to prevent 21 It generally takes 12 to 14 weeks to train new public assistance specialists on the FLORIDA system. Added training is needed when policy or other changes require different client information or other application processes. 9 benefit overpayments and to recover payments that have been paid in error. In carrying out this responsibility, the department uses a number of front-end fraud prevention techniques to verify client information. 22 The department also uses a benefit recovery process that assesses cases that may have resulted in overpayments and arranges to recover identified overpayments or refers potential fraud cases to the Florida Department of Law Enforcement for investigation. To assess the efficiency of these efforts, the Legislature has required the department to report the return on investment (ROI) for these activities since 2001-02. 23 Because the department had not calculated or reported ROI as required, we worked with department managers to develop a valid methodology to calculate the ROI for 2000-01 through 2002-03 for each unit and for the effort as a whole. Front-end fraud prevention yields higher returns than benefit recovery. As shown in Exhibit 7, the overall ROI for fraud prevention and benefit recovery activities was about $2 for each $1 invested in each of the past three years. In addition, the program realizes a much higher return from front-end fraud prevention efforts than from benefit recovery efforts. This is not unexpected since the department has more control over preventing overpayments than in recovering funds from clients who may have limited capacity to repay. However, for the past two years the department spent more to support benefit recovery activities than it recovered. 24 The department should explore ways to improve the efficiency of its benefit recovery activities to improve efficiency to reach the break-even point. For example, the department could examine its contract with the Florida Department of Law Enforcement, since it represents half of the expenditures for benefit recovery activities. 22 See Appendix D for a detailed description of front-end fraud prevention and benefit recovery procedures. 23 Return on investment refers to the ratio of dollars collected to dollars spent. 24 See Appendix D for detailed expenditure and recovery information and ROI calculations.

Exhibit 7 Front-End Fraud Prevention Efforts Yield Greater Returns Than Benefit Recovery Return on Investment Fiscal Year Activity 2000-01 2001-02 2002-03 Front-End Fraud Prevention 1 $6.85 $7.06 $6.49 Benefit Recovery 1 1.01 0.93 0.92 Total $2.07 $2.03 $2.00 1 Front-end fraud prevention returns are estimates of the costs avoided by the state for not paying benefits for which clients are not entitled. Benefit recovery returns are actual dollars collected. Source: Department of Children and Families. Although the Legislature currently requires an overall ROI that includes both front-end fraud prevention and benefit recovery, we believe the Legislature should also require the department to report separate ROIs for front-end fraud prevention and benefit recovery. Based on the above results, the Legislature should increase the overall ROI standard from the current level of $0.84 to at least $2. Further, the Legislature should set the ROI for benefit recovery at $1 (the break-even point) and the ROI for front-end fraud prevention at a minimum of $6.50. This would help spur the program to improve its performance and maximize prevention and recovery activities of public assistance fraud. The department has reduced the benefit recovery backlog to its lowest level in years. In recent years, three separate federal and state reviews raised concerns regarding the ESS benefit recovery process. 25 These reports found that since the early 1990s the program had not investigated cases referred for recovery of overpayments in a timely manner, resulting in significant backlogs. 26 The backlog was so severe and had existed for so long that the 2001 USDA inspector general report recommended that the USDA withhold federal administrative funds if the department did not correct backlog and other problems. To avoid losing administrative funds, the department initiated steps to address this problem and by June 2003 reduced the backlog from a reported high of 380,000 to fewer than 25,000 cases. 27 USDA food stamp administrators have reviewed these efforts and indicated that the department has reasonably addressed the backlog and is not currently at risk of losing administrative funds. The department is taking several steps to prevent benefit recovery backlogs from growing in the future. These efforts include enhancing the FLORIDA system to automatically forward preliminary referrals from districts to the benefit recovery central office or to FDLE after 10 days, better managing cases among districts by pairing low volume districts with high volume districts to more evenly distribute workload, and improving training to place additional emphasis on how to recognize cases that should be referred for benefit recovery action. However, the department is no longer actively monitoring benefit recovery backlogs because the federal government is satisfied with the current status. Without ongoing monitoring, a historically significant backlog problem could reemerge. To maintain backlogs at reasonable levels, the department should develop a system to monitor backlogs. This system should alert department administrators when the number of backlog cases reaches a level that would require more than three to six months effort to resolve. Given current resources, a six-month backlog 25 Management of the Benefit Overissuance Recovery Process by the Department of Children and Family Services, State of Florida Auditor General, Report No. 13591, February 2000. Review of Benef it Recovery Backlog for the Period June 7, 1998 Through June 7, 2001, Florida Department of Children and Families Office of the Inspector General, Management Review#02-05-M. September 2001 Food and Nutrition Service Florida Food Stamp Program, USDA Office of Inspector General Southwest Region, Report No. 27004-3-AT, November 2001. 26 A case is classified as backlog when its referral date is more than 90 days old, and the case has not been investigated. 10 27 The majority of this reduction (185,000 cases) was eliminated under a one-time agreement between the department and the USDA to write off old and low value cases. The remaining reduction is attributed to the department hiring a consultant to help prioritize cases and assist in developing an automated review process. The department then used internal workgroups to review backlog cases, identify cases that could be closed, and prioritize the remaining cases for processing.

would equal between 25,000 and 50,000 cases. 28 This would help ensure that the department continues to pursue public assistance recovery activities in a timely manner when it detects overpayments. Conclusions and Recommendations The ESS Program has improved timeliness in processing applications for public assistance benefits over the last two years. However, it has not reported TANF accuracy to the Legislature as required, and its food stamp error rates exceed the national average, which has resulted in federal financial penalties totaling more than $2.2 million for federal Fiscal Years 2001 and 2002. The department s front-end fraud prevention and benefit recovery efforts have realized a return of about $2 for each $1 invested over the past three years. However, fraud prevention efforts have produced a much larger return on investment than recovery efforts, which have not recovered expenses in the past two years. Although benefit recovery backlogs are currently relatively low, the department has experienced huge backlogs in the past resulting in overpayments that were never collected. To improve the ESS Program, we recommend that the Legislature direct the department to take the actions described below. Resume tracking and reporting TANF accuracy rates. In addition to being required by the Legislature as part of the PB 2 process, tracking accuracy is critical to ensuring that TANF block grant monies are spent wisely and are available to meet the needs of citizens. 28 The conversion of 50,000 cases into a six-month backlog is based on the average number of reviews completed on a monthly basis per investigator from October 1, 2002, through June 30, 2003. If program resources (i.e., the number of staff) increase or decrease, the actual number of cases considered a six-month backlog will change accordingly. 11 Reduce food stamp payment errors. To avoid future financial penalties, it is critical that the department reduce food stamp payment errors. If the state s error rate equals or falls below the national average in federal Fiscal Years 2004 and 2005, the USDA will forgive half of the most recently assessed $1.7 million penalty. To reduce food stamp errors to this level will require a multifaceted approach. In addition to actions already underway, the department should set quarterly targets for each district and region, develop training for public assistance workers focused on assisting them to understand and apply policies to reduce typical errors, and evaluate activities such as change units and make decisions to expand or adjust activities based on evaluation results. The department should annually report to the Legislature, with the first report due by December 31, 2003, on its efforts to evaluate these activities and their effect on reducing error rates. Develop an action plan focused on pursuing federal performance bonuses under the 2002 Farm Bill. The department should compete for a share of the $48 million the USDA will award to states for high or improved performance. While specific criteria may change in federal Fiscal Year 2004, the USDA will award bonuses for federal Fiscal Year 2003 based on application processing timeliness, payment accuracy, low negative error rates, and high participation rates. Require the department to report separate returns on investment (ROIs) for front-end fraud prevention and benefit recovery. Depending on legislative direction, the department could report these ROIs instead of or in addition to the currently required combined ROI for these activities. Performance standards for these measures should be set at no less than $1 (the breakeven point) for benefit recovery, a minimum of $6.50 for front-end fraud prevention, and at least $2 for the two efforts combined.

Develop a system to monitor benefit recovery backlogs. This is important to ensure that backlogs do not become high enough to risk losing federal administrative funds. The monitoring system should alert the department when the backlog reaches a level requiring three to six months of resources to resolve. In that event, the department should redeploy resources to reduce the backlog to current levels. Agency Response Pursuant to s. 11.51(5), Florida Statutes, the Department of Children and Families was provided a draft copy of our report. The inspector general s response is on the following pages. 12

Appendi x A Justification Review Summary Program Evaluation and Justification Review Conclusions Section 11.513(3), Florida Statutes, provides that OPPAGA Program Evaluation and Justification Reviews shall address nine issue areas. Our conclusions on these issues as they relate to the Economic Self- Sufficiency Program in the Department of Children and Families are summarized below. Table A-1 Summary of the Program Evaluation and Justification Review of the Economic Self-Sufficiency Program Issue The identifiable cost of the program The specific purpose of the program, as well as the specific public benefit derived therefrom Progress towards achieving the outputs and outcomes associated with the program An explanation of circumstances contributing to the state agency s ability to achieve, not achieve, or exceed its projected outputs and outcomes, as defined in s. 216.011, F.S., associated with the program OPPAGA Conclusions For Fiscal Year 2003-04, the Legislature appropriated $661.1 million for the Economic Self-Sufficiency Program. This includes $395.5 million in General Revenue or 59.8 % of the total appropriation. The program spends 93.9% of its appropriations providing direct services and benefits to low income individuals and families in Florida, including refugees. Florida s Economic Self-Sufficiency Program assists low-income individuals become and remain stable and selfsufficient. The program s benefits help individuals and families obtain basic necessities, such as food, housing, and medical care for themselves and their families. The program determines applicant eligibility for a number of state and federal public assistance benefits including TANF cash assistance, Food Stamps, Medicaid, Optional State Supplementation funds and refugee cash assistance. During Fiscal Year 2002-03, on average, the program provided food stamp benefits to 1,024,692 individuals and provided TANF cash assistance to 129,402 individuals, on average, each month. During that same year, the program determined Medicaid eligibility for 2,304,769 individuals. In Fiscal Year 2002-2003, the Economic Self Sufficiency Program essentially met the overall timeliness standard (99%) for processing public assistance applications, processing 98.97% of the application within mandated timeframes. This represents significant and continued improvement over 2000-01 performance (95.8%). While falling short of meeting the 98% standard for processing Optional State Supplementation (OSS) applications within 90 days, the department s 2002-03 performance (95%) represents significant and continued improvement over 2000-01 performance (75.6%) and 2001-02 performance (90.8%). Although required by as part of the PB 2 process, the department ceased monitoring TANF payment error rates at the end of federal Fiscal Year 2001. The payment error rate for that year was 6.23%, resulting in overpaying clients by $13.6 million. Florida s food stamp payment error rate remained above the national average in federal Fiscal Year 2002, resulting in the USDA imposing a financial penalty of over $1.7 million on the state. The payment error rate was 9.61%, resulting in overpaying clients $65.2 million in excess benefits and underpaying $19.2 million in benefits to eligible clients. In 2002-03, the program s overall return on investment for its front-end fraud prevention and benefit recovery activities was $2. Front-end fraud prevention yielded higher returns than benefit recovery, realizing nearly $6.50 in costs avoided for every $1 invested while benefit recovery collected $0.92 for each $1 invested. Program administrators have stated that increased applications, unstable caseloads, and frequent changes in policy all contribute to the program s ability to achieve or not achieve expected outcomes. Most notably, time limits imposed by welfare reform as well as changing economic conditions in the state have resulted in a changing client population and policy environment that can directly affect public assistance workers ability to process applications timely and determine benefits accurately. 13

Issue Alternative courses of action that would result in administering the program more efficiently and effectively The consequences of discontinuing the program Determination as to public policy; which may include recommendations as to whether it would be sound public policy to continue or discontinue funding the program, either in whole or part Whether the information reported pursuant to s. 216.03(5), F.S., has relevance and utility for the evaluation of the program Whether state agency management has established control systems sufficient to ensure that performance data are maintained and supported by state agency records and accurately presented in state agency performance reports Source: OPPAGA. OPPAGA Conclusions The following summarizes the alternative courses of actions recommended to the Economic Self-Sufficiency Program by OPPAGA. Implementation of these recommendations should improve the efficiency and effectiveness of the state s Economic Self-Sufficiency Program. Track and report TANF accuracy rates. In addition to being required by the Legislature as part of the PB 2 process, tracking accuracy is critical to ensuring that TANF block grant monies are spent wisely and are available to meet the needs of citizens. Reduce food stamp payment errors. To avoid future financial penalties, it is critical that the department reduce food stamp payment errors. If the state s error rate equals or falls below the national average in federal Fiscal Years 2004 and 2005, the USDA will forgive half of the most recently assessed $1.7 million penalty. Reducing food stamp errors to this level will require a multi-faceted approach. In addition to actions already underway, the department should set quarterly targets for each district and region, developing training for public assistance workers focused on assisting them to understand and apply policies to reduce typical error, evaluating activities such as change units and make decisions related to expanding or adjusting activities based on evaluation results. The department should annually report to the Legislature, with the first report due by December 31, 2003, on its efforts to evaluate these activities and their effect on reducing error rates. Develop an action plan focused on pursuing federal performance bonuses under the 2002 Farm Bill. The department should compete for a share of the $48 million the USDA will award to states for high or improved performance. While specific criteria may change in federal Fiscal Year 2004, the USDA will award bonuses for federal Fiscal Year 2003 for application processing timeliness, payment accuracy, low negative error rates, and high participation rates. Require the department to report separate ROIs for front-end fraud prevention and benefit recovery. Depending on legislative direction, the department could report these ROIs instead of or in addition to the currently required combined ROI for these activities. Performance standards for these measures should be set at no less than $1 (the break-even point) for benefit recovery, a minimum of $6.50 for front-end fraud prevention, and at least $2 for the two efforts combined. Develop a system to monitor benefit recovery backlogs. This is important to ensure that backlogs do not become high enough to risk losing federal administrative funds. The monitoring system should alert the department when the backlog reaches a level requiring three to six months of resources to resolve. In that event, the department should redeploy resources to reduce the backlog to current levels. Florida s Economic Self-Sufficiency Program assists families to become and remain stable and self-sufficient. Discontinuing the program would increase the number of Floridians without access to basic supports, food, and shelter during times of personal economic crisis and instability. In addition, discontinuing the program could result in loss of the state s ability to receive federal funding ($1,209,296,593 issued in Fiscal Year 2002-03 for food stamps, TANF, refugee assistance and the federal share of OSS payments) to help pay for basic supports needed for families and individuals with limited resources. The Economic Self-Sufficiency Program provides beneficial services to clients and is a cost benefit to taxpayers. The report identifies strategies that could improve program effectiveness and efficiency. The Economic Self-Sufficiency Program s outcome measures directly address the program s major responsibility to provide timely and accurate public assistance benefits to eligible families and individuals. The program s timeliness measures assess whether the program is determining eligibility and providing benefits to families and individuals who are experiencing financial difficulties within reasonable timeframes. Payment accuracy measures provide an indication of the state s stewardship of both federal and state funds. When accuracy is low, the state can not ensure that sufficient funds are available to provide benefits to eligible citizens and can result in federal sanctions. Return on investment is a good indicator of program efficiency and should be reported for both frontend fraud prevention and benefit recovery activities. The Department of Children and Families Office of Inspector General completed an assessment of the Economic Self-Sufficiency s performance data in June 2003. This review examined the timeliness of performance data in four districts. Generally, the review found that requests for assistance were frequently absent from client files or not date stamped, limiting the ability of the reviewers to verify whether standards were met. However, data from the FLORIDA, (Florida On-Line Recipient Information Data Access) system has been validated and is periodically reviewed by the Department of Health and Human Services and the United States Department of Agriculture, the two primary federal agencies providing funds for benefits and services offered by the Economic Self-Sufficiency Program. 14

Appendi x B Food Stamp Incentive Funding Prior to federal Fiscal Year 2003, the federal government provided states with incentive funding if they achieved food stamp error rates below 6%. Specifically the USDA awarded states an additional 1% of their food stamp administrative costs for each 0.1% reduction in error rate below 6%. Florida did not earn any incentive funding during this time; in fact, Florida s lowest error rate during this period was 9.4%. In addition, Florida s error rates ranked in the bottom third of all states and territories in four of the last five federal fiscal years. Florida also ranked poorly compared to other large states. Table B-1 shows the amount of enhanced funding earned by 16 states and the Virgin Islands for achieving rates below 6%. Texas, Wyoming, and South Dakota received enhanced funding in each of these years. Table B-1 States Received $250.7 Million in Federal Enhancement Funds From Federal Fiscal Year 1998 Through Federal Fiscal Year 2002 State Error Rate Amount Received Federal Fiscal Year 1998 Texas Minnesota Hawaii South Dakota Wyoming Arizona 5.27 5.18 4.82 2.11 4.81 5.90 $ 19,742,234 4,494,636 1,700,458 652,879 557,462 298,096 Total $ 27,445,765 Federal Fiscal Year 1999 Texas Mississippi Arkansas South Carolina South Dakota Wyoming 4.56 4.91 4.54 5.79 2.19 2.91 $ 27,941,372 5,161,155 4,099,364 757,997 714,305 548,871 Total $ 39,223,064 Federal Fiscal Year 2000 Texas Minnesota Mississippi Arkansas South Carolina Louisiana West Virginia Tennessee South Dakota Wyoming Kentucky 4.14 3.58 4.69 4.03 4.47 5.66 5.09 5.71 1.18 4.01 5.81 $ 28,660,790 6,603,094 4,958,828 4,002,976 3,725,822 2,012,364 1,897,508 1,259,171 731,693 622,771 595,110 Total $ 55,070,127 State Error Rate Amount Received Federal Fiscal Year 2001 Texas Mississippi Minnesota South Carolina Arkansas Louisiana South Dakota Wyoming Virgin Islands Rhode Island 3.73 3.47 5.22 4.62 3.24 5.78 2.11 3.04 4.70 5.56 $ 29,856,741 4,775,951 4,680,835 4,392,948 4,312,274 1,305,312 768,456 678,387 519,408 470,286 Total $ 51,760,598 Federal Fiscal Year 2002 Texas New Jersey North Carolina Mississippi South Carolina Arkansas Arizona Hawaii Louisiana Minnesota Wyoming South Dakota Virgin Islands 4.85 4.08 4.70 4.39 4.40 4.29 5.27 5.03 5.78 5.73 3.29 2.12 5.72 $ 29,136,689 14,452,563 10,122,355 4,898,620 4,865,345 3,967,618 3,841,799 1,476,790 1,443,356 1,423,066 750,857 728,325 124,640 Total $ 77,232,023 Source: USDA Food and Nutrition Service. 15

Appendix C Federal Food Stamp Options and Waivers Options and waivers simplify rules, regulations and practices associated with eligibility determination and recertification, and thus can help improve service and decrease errors. To use an option, states need only submit a letter of intent to the USDA. However, before a state can use a waiver, it must be approved by the USDA. The state must submit a formal application that outlines the state s needs and expected benefits for the particular waiver. Florida is currently taking advantage of four options and seven waivers. ESS administrators also are pursuing or evaluating three additional options and two additional waivers. The program is not actively pursuing three available federal options either because a pending waiver would have the same effect or because the state does not have enough recipients who fit criteria to justify the administrative expense of using the option. Table C-1 describes these options and waivers in more detail. (This listing is not inclusive of all federally available waivers.) Table C-1 Florida Is Using Most Available Options and Is Pursuing Additional Waivers Options Currently Used in Florida States can exclude some resources that are also excluded by TANF or Medicaid Simplified utility allowance Allows standard deductions for heating/cooling of $198 In the past, if several families lived together, they split the standard deduction, with each only claiming half. The option allows each family group to claim the full deduction. Simplified determination of housing Allows homeless individuals/families to claim a credit of $143 per month for housing Status reporting earned income A household with earned income can certify for up to six months, only reporting changes in income status (e.g., move from part time to full time, change in pay rate or lost job). The household would not need to report fluctuations in income. Options Being Pursued or Evaluated in Florida Child support Excludes legally obligated child support from income. Treats legally obligated child support to a non-household person as an income exclusion instead of a deduction. Options Not Being Pursued in Florida Transitional benefits: Families leaving cash assistance can continue receiving food stamps. Status The department chose to combine the food stamp and TANF polices through an option offered in 2000. The policy is to exclude as many resources as available. The department pursued this option, and it was effective April 1, 2003. The department has been using this option since June 1, 2001. The department has been using this option since 2001. Florida is evaluating this option. The state has no plans to implement this option, believing that it poses a risk to the state, with a high potential for errors, which must be recouped; errors would occur because other changes to a household s financial status would still need to be acted upon even though food stamps remain unchanged; it would be especially difficult to implement since all 16

Simplified determination of deduction Allows household to defer reporting changes and seek early recertification visits for some types of changes in income Excludes changes in housing allowances and earned income Other income exclusions Can exclude other income from food stamp eligibility determination that is also excluded in cash assistance and Medicaid Waivers Being Used in Florida Required second interview for food stamp benefits Time limit of 3 out of 36 months to receive food stamps and work requirements for able-bodied adults, aged 8-24 years, without dependents Elder Simplified Application Process Status reporting Florida has pursued a waiver to allow greater flexibility pertaining to status reporting. The waiver allows families to defer reporting changes for six months, including changes in earnings, unless income changes exceed 130% of the original income. Under $10 reporting exclusion COAP Combined Outreach and Application Project Source: Department of Children and Families, 2003. assistance programs are integrated into one computer system, and would lead to changes to food stamps that should not occur; and it could be confusing for households with different reporting requirements across programs. This option is not being pursued in Florida since a waiver is currently under consideration by the federal government. Florida is not pursuing this option. There is not enough of an impact to warrant pursing, since it is limited to items like some types of education aid and does not affect many applicants. Status Florida can waive a second interview for food stamp eligibility when the certification period is short and the household also receives cash assistance; for example, if eligible for 30-45 days of cash assistance, state can match the certification period for cash and food stamps. The state has two waivers affecting this category. A state waiver eliminated the time limit for food stamp eligibility for individuals in this category. Remote counties or those without work opportunities or public transport do no have same work requirements. This waiver reduces certain application requirements for seniors. Earned income: A household with earned income can certify for up to six months, only reporting changes in income status (e.g., move from part-time to full-time, change in pay rate or lose job). The household would not need to report fluctuations in income. This waiver was approved in July 2003. This removes the requirement that recipients self-report interest income that is less than $10. This allows states to merge the food stamp and TANF application with the federal SSI application process. If this occurs, any applicants inappropriately approved for benefits or continued benefits are not considered an error attributable to the state, but are considered the responsibility of the federal program. This would decrease some administrative processing and would have a direct effect of reducing the states error rate. The department s letter of intent and overview has been approved, with implementation set for October 2004. Until then the department will develop all steps needed for full implementation. 17

Appendix D Front-End Fraud Prevention and Benefit Recovery Front-End Fraud Prevention Improves Payment Accuracy Front-end fraud prevention improves payment accuracy by using various techniques to verify client information. These techniques help ensure that eligibility is determined quickly and increase the certainty that benefits are determined accurately. When clients apply or reapply for benefits, public assistance workers attempt to verify client information using readily available data. The three primary sources of data include Unemployment Compensation Benefits data system; employment information available through Data Exchange, an employment database; and the Family Assistance Information Report, a public assistance database system. When client information cannot be quickly verified, the public assistance worker refers the case to the Front-end fraud prevention unit to confirm the information using additional resources. These resources include criminal justice files to verify past incarceration, payroll data for current income, motor vehicle data for asset determination, and information related to bankruptcies, liens, and use of fictitious names for business operations. Front-end fraud employees also interview neighbors and employers to verify hours and pay for current employment, the number of household members, and other personal data. Most eligibility problems result from unintentional client error or public assistance worker s misunderstanding. If a case has clear and convincing evidence that incorrect information was intentionally provided, it is referred for an administrative disqualification hearing. For all cases, based on the additional information, the front-end fraud prevention unit will approve or deny eligibility for assistance. Table D-1 gives the amount of expenditures for front-end fraud prevention activities for the last three fiscal years as well as the costs avoided as a result of these activities. The exhibit also gives the return on investment for these activities. Based on this three-year period, front-end fraud expenditures have remained stable with the unit spending on average $3,081,838 per year and achieving cost avoidance of $20,940,689 on average per year. The combination of low expenditures and high levels of cost avoidance has resulted in returns averaging a $6.81 return for every dollar spent. Table D-1 Front-End Fraud Prevention Activities Returned Between $6.49 and $7.06 Over the Last Three Years Front-End Fraud Prevention Return on Investment Fiscal Year Expenditures Cost Avoidance Return on Investment 2000-01 $3,056,591 $20,950,985 6.85 2001-02 3,024,462 21,345,838 7.06 2002-03 3,164,463 20,525,244 6.49 Source: Department of Children and Families. Benefit Recovery Recoups Overpayments The benefit recovery unit recoups overpayments due to either agency or client error. The USDA 18

requires states to implement benefit recovery activities and allows states to retain up to 35% of food stamp recovered payments due to client error. However, all recovered overpayments that are due to errors made by public assistance workers must be returned to the USDA. Benefit recovery employees review and verify referral information, assess whether overpayments have occurred, determine the amount of overpayments, arrange to collect overpaid funds, and refer suspected fraud cases to the Florida Department of Law Enforcement for investigation and possible prosecution. Clients currently receiving benefits can repay funds by reducing benefits until overpayments are recouped or through a scheduled repayment plan. If clients are no longer receiving benefits, the department can recover overpayments by having clients develop repayment plans or initiate payroll deductions, intercepting federal income tax refunds, or placing liens on client property. When scheduled repayments become delinquent, employees refer the case to a contracted collection agency. Table D-2 displays benefit recovery expenditures and collections for the last three fiscal years. On average, the benefit recovery unit spent $13,556,238 and collected $12,897,779 per year. While the return on investment is significantly lower for benefit recovery activities than for front-end fraud prevention, it is important to remember that the department has more control over preventing overpayments than in recovering overpayments from clients who may have limited capacity to repay. The low return on investment indicates that benefit recovery administrators should look for ways to become more efficient. Ideally, the unit s return on investment should not move below 1.00 (the break-even point). Table D-2 Return on Investment for Benefit Recovery Was Below 1.00 in 2001-02 and 2002-03, Which Calls for Activities to Improve Efficiency Benefit Recovery Return on Investment Fiscal Year Expenditures Cost Avoidance Return on Investment 2000-01 $13,750,904 $13,895,925 1.01 2001-02 13,822,365 12,793,617 0.93 2002-03 13,095,446 12,003,795 0.92 Source: Department of Children and Families. 19

The Florida Legislature Office of Program Policy Analysis and Government Accountability Visit the Florida Monitor, OPPAGA s online service. See http://www.oppaga.state.fl.us. This site monitors the performance and accountability of Florida government by making OPPAGA's four primary products available online. OPPAGA publications and contracted reviews, such as policy analyses and performance reviews, assess the efficiency and effectiveness of state policies and programs and recommend improvements for Florida government. Performance-based program budgeting (PB²) reports and information offer a variety of tools. Program evaluation and justification reviews assess state programs operating under performance-based program budgeting. Also offered are performance measures information and our assessments of measures. Florida Government Accountability Report (FGAR) is an Internet encyclopedia of Florida state government. FGAR offers concise information about state programs, policy issues, and performance. Best Financial Management Practices Reviews of Florida school districts. In accordance with the Sharpening the Pencil Act, OPPAGA and the Auditor General jointly conduct reviews to determine if a school district is using best financial management practices to help school districts meet the challenge of educating their students in a cost-efficient manner. Subscribe to OPPAGA s electronic newsletter, Florida Monitor Weekly, a free source for brief e-mail announcements of research reports, conferences, and other resources of interest for Florida's policy research and program evaluation community. OPPAGA provides objective, independent, professional analyses of state policies and services to assist the Florida Legislature in decision making, to ensure government accountability, and to recommend the best use of public resources. This project was conducted in accordance with applicable evaluation standards. Copies of this report in print or alternate accessible format may be obtained by telephone (850/488-0021 or 800/531-2477), by FAX (850/487-3804), in person, or by mail (OPPAGA Report Production, Claude Pepper Building, Room 312, 111 W. Madison St., Tallahassee, FL 32399-1475). Florida Monitor: http://www.oppaga.state.fl.us/ Project supervised by Yvonne Bigos, Chief Legislative Analyst (850/487-9230) Project conducted by Michael Garner (850/487-9252) and Rae Hendlin (850/410-4795) Frank Alvarez (850/487-9274) Staff Director Gary R. VanLandingham, OPPAGA Interim Director 20